Why Is My Billionaire Coming Up Short with Affluence Insight Results?

Carol Belair

If you are a Target Analytics client who has an Affluence Insight subscription and you’ve started reviewing the results, this blog might be helpful to having a deeper understanding of the data you have at your fingertips.  First step we recommend is to check out my colleague Dave Lamb’s very informative blog post on the Community, “Where do Target Analytics Affluence values come from?”  Also we have a fantastic new resource via the Blackbaud Institute entitled, Supporters In Sight Part 2: A Look at Affluent Donor Personas.”  These are both helpful resources that we highly recommend our clients look at when thinking about utilizing this information.

For this blog post, I worked with Dave to provide a deeper understanding of Affluence Insight data – in particular as it comes to understanding results for the four Wealth Attributes – as it pertains to the ultra-wealthy billionaires as well as how to best apply this type of data to your various fundraising programs.

Each person’s Affluence score is on a continuum of accuracy from right on the mark – to slightly off the mark – to very far off the mark.  These attributes are generated by models that use consumer marketing data as their primary sources.  Models deal with probability so like any model, there will be some variability in how accurately they predict what they are trying to predict.  If you plot actual income on a chart for those whose income is predicted to be $100K , the plot points will scatter around the $100K mark in some version of a bell curve.  Some people whose income is predicted at $100K will in reality, only have an income of $60K, $70K or $80K.  Others will have a real income of $120K or more, but most of the results for annual income will be within one standard deviation from the predicted value. 

This technology grows out of the marketing world.  If you look at it from the standpoint of a marketer, the goal is to obtain a list of people who mostly qualify to make a purchase, investment, etc.  If some people selected don’t qualify, this is an acceptable result as long as most people selected do. 

To look at it differently, if you pick any person from a nonprofit database at random, your odds of picking one with a 6-figure income will be very poor.  But if you filter the database to those whose Affluence income is $100K+, your odds improve considerably.  You can be almost 100% sure that the attribute is never going to be precisely right, but you can have confidence that most people’s scores put them in the right wealth category.  Then for some people that score will be off by quite a bit, but it is not a large proportion of the population.

These attributes represent statistically derived estimates, so it is not surprising that verified billionaires would have lower Affluence results than expected.  The people on the Forbes 400 spend their money and hold their assets in ways that are quite different from the ordinary wealthy person.  They often have intermediaries who manage their finances and purchases for them such that the consumer marketing data do not represent the reality of their finances well.

Affluence Insight does not replace the detailed publicly available data that is available via a WealthPoint search in ResearchPoint.  Think of it as a compliment to this information and a great way to segment your database.  I find that many clients I work with find the most useful application to be in utilizing the Wealth Segmentation Donor Categories (A1, A2 or A3 Philanthropists, B1, B2 or B3 Humanitarians, etc.) with messaging opportunities for their donor population and also a way to further refine prioritizing prospects within custom modeling results for similar scoring prospects for discovery opportunities.

One quick reminder that if you are a Raiser’s Edge client and have not imported your Affluence Insight results into Raiser’s Edge, be sure that you check out this blog post I published back on October 31, 2018, entitled, “Synchronizing Affluence Insight™ Via RP-RE Integration” first.  The blog walks you through a step-by-step process for configuring the integration so that the Affluence Insight scores  import into Raiser’s Edge.  If you do not configure these fields in ResearchPoint first, the results will not import into Raiser’s Edge and you’ll have to do it over again.  My colleague Ryan Corder also commented on this blog via a comment at “Thoughts on integrating Affluence Insight data with Raiser's Edge”.  His comment contains more information on the process of creating a Research List in ResearchPoint that will allow you to import the Affluence Insight scores via integration.